ImmersiveAmbientModecolor: #c89d97 (color 1)
Video Format : 22 (720p) openh264 ( https://github.com/cisco/openh264) mp4a.40.2 | 44100Hz
Audio Format: Opus - Normalized audio
PokeTubeEncryptID: 8bb281b03bd6f65f06327c29c402496358c11251d8a1d598e6c080cb1bb2f8581093de633825fa173c0eeba272ddcada
Proxy : eu-proxy.poketube.fun - refresh the page to change the proxy location
Date : 1716044229874 - unknown on Apple WebKit
Mystery text : NW5uU1NqZXI0ajAgaSAgbG92ICB1IGV1LXByb3h5LnBva2V0dWJlLmZ1bg==
143 : true
9,734 Views β’ Apr 21, 2018 β’ Click to toggle off description
"I'm excited about understanding documents. Instead of giving Google some words and it will find you some documents that have all those words, you give Google some themes and it will find you some documents that are conveying those themes even if there's using different words." - Geoffrey Hinton (2016)
YouTube playlist: bit.ly/2K4rhh9
Full: bit.ly/2HhS47D
CONNECT:
- AI Podcast: lexfridman.com/ai/
- Subscribe to this YouTube channel
- LinkedIn: www.linkedin.com/in/lexfridman
- Twitter: twitter.com/lexfridman
- Facebook: www.facebook.com/lexfridman
- Instagram: www.instagram.com/lexfridman
- Slack: deep-mit-slack.herokuapp.com/
Views : 9,734
Genre: Science & Technology
Date of upload: Apr 21, 2018 ^^
Rating : 4.875 (6/186 LTDR)
RYD date created : 2022-01-24T02:43:44.903666Z
See in json
YouTube Comments - 6 Comments
Top Comments of this video!! :3
Intriguing--I'm trying to marry pencil-and-paper linguistics to AI insofar as it helps facilitate a better understanding of the link between language and consciousness, particularly sentience. If anyone can help point me in the right direction, cheers!
1 |
"That would be kinda useful."
4 |
Commercial product: https://catalog.m-files.com/shop/m-files-smart-subjects/
|
Is this almost a prediction of chatGPT/GPT4?
|
@myothersoul1953
6 years ago
I agree with him, machine translation will improve and we will be able to do searches using terms, phrases that have the same meaning but are not composed of the same words. However I don't think that means the AI understands anything, it will recognized patterns in the many thousands of examples it is trained on but that means it understands the meaning. I'll even stick my neck out and make a more precise prediction. My hypothesis is when the AI's errors are examined there will be more semantic mistakes than grammatical mistakes.
1 |